HYDROXYLATED POLYBROMINATED DIPHENYL ETHERS IN...

90
HYDROXYLATED POLYBROMINATED DIPHENYL ETHERS IN BALTIC SEA BIOTA Natural production, food web distribution and biotransformation Dennis Lindqvist

Transcript of HYDROXYLATED POLYBROMINATED DIPHENYL ETHERS IN...

  • H Y D R O X Y L A T E D P O L Y B R O M I N A T E D D I P H E N Y L E T H E R S I N B A L T I C S E A B I O T A N a t u r a l p r o d u c t i o n , f o o d w e b d i s t r i b u t i o n a n d b i o t r a n s f o r m a t i o n

    Dennis Lindqvist

  • Hydroxylated polybrominated diphenyl ethers in Baltic Sea biota

    Natural production, food web distribution and biotransformation

    Dennis Lindqvist

  • ©Dennis Lindqvist, Stockholm University 2016 ISBN Printed 978-91-7649-585-8 ISBN PDF 978-91-7649-586-5 Printed in Sweden by US-AB, Stockholm 2016 Distributor: Department of Environmental Science and Analytical Chemistry, Stockholm University

  • “When you can’t run anymore, you crawl, and when you can’t do that, you find someone to carry you”

    -Joss Whedon’s Firefly ep.12-

    Till alla er som hjälpt till att bära

  • i

    List of papers

    This thesis is based on the following papers, referred to in the text by their roman numerals, I-IV:

    I Lindqvist D., Dahlgren E., and Asplund L. (2017) Biosynthesis of hydroxylated polybrominated diphenyl ethers and the correlation with photosynthetic pigments in the red alga Ceramium tenuicorne. Phyto-chemistry 133, 51-58.

    II Lindqvist D., Jensen S., and Asplund L. (2014). Lipid-soluble conju-gates of hydroxylated polybrominated diphenyl ethers in blue mussels from the Baltic Sea. Environ. Sci. Pollut. Res. 21, 954-961.

    III Dahlgren E., Lindqvist D., Dahlgren H., Asplund L., and Lehtilä K. (2016) Trophic transfer of naturally produced brominated aromatic compounds in a Baltic Sea food chain. Chemosphere 144, 1597-1604.

    IV Lindqvist D. and Asplund L. (Manuscript) Determination of hydrox-ylated polybrominated diphenyl ethers in blood from Baltic grey seals.

    All published material is reprinted with the permission of the publishers.

    The following papers have also contributed to the content of this thesis but are not included:

    [1] Jensen S., Lindqvist D., and Asplund L. (2009). Lipid extraction and determination of halogenated phenols and alkylphenols as their pen-tafluorobenzoyl derivatives in marine organisms. J. Agric. Food Chem. 7, 5872–5877.

    [2] Löfstrand K., Liu X., Lindqvist D., Jensen S., and Asplund L. (2011). Seasonal variation of hydroxylated and methoxylated brominated di-phenyl ethers in blue mussels from the Baltic Sea. Chemosphere 84, 527-532.

    [3] Dahlgren E., Enhus C., Lindqvist D., Eklund B., and Asplund L. (2015). Induced production of brominated aromatic compounds in the alga Ceramium tenuicorne. Environ. Sci. Pollut. Res. 22, 18107-18114.

  • ii

    Table of contents

    List of papers .................................................................................................. i Table of contents ...........................................................................................ii Abbreviations ............................................................................................... iv 1 Background ................................................................................................ 1

    1.1 Aims of the thesis ................................................................................. 3

    2 HPC essentials ............................................................................................ 4 2.1 Natural production of HPCs ................................................................. 4

    2.1.1 Biosynthesis of bromophenols (BPs) ............................................ 5 2.1.2 BP dimerization and production of OH-PBDEs ........................... 7 2.1.3 Benefit of the production .............................................................. 8

    2.2 Anthropogenic sources of HPCs .......................................................... 9 2.2.1 Intentional and accidental releases .............................................. 9 2.2.2 Metabolites of anthropogenic organohalogen pollutants ........... 10

    2.3 HPCs, physicochemical properties and kinetics ................................. 11 2.3.1 Physicochemical properties ........................................................ 11 2.3.2 Adsorption, distribution, metabolism, excretion (ADME) .......... 11

    2.4 HPC toxicity, with focus on OH-PBDEs ........................................... 13 2.4.1 Endocrine disruption .................................................................. 13 2.4.2 OXPHOS disruption ................................................................... 14

    2.5 Ecosystem concerns and the larger perspective ................................. 16 2.5.1 Xenobiotic exposure and metabolic disruption .......................... 16 2.5.2 Co-affecting factors, thiamine deficiency ................................... 17

    3 Samples and sample collection ................................................................ 19 3.1 Species studied in this thesis .............................................................. 19

    3.1.1 Ceramium tenuicorne ................................................................. 19 3.1.2 Blue mussel ................................................................................. 20 3.1.3 Gammarus spp. ........................................................................... 20 3.1.4 Three-spined stickleback ............................................................ 21 3.1.5 Perch ........................................................................................... 22 3.1.6 Grey seal ..................................................................................... 22

    3.2 Sample collection ............................................................................... 23

    4 Analytical methods................................................................................... 25 4.1 Extraction ........................................................................................... 25

    4.1.1 The lineage of the Jensen methods ............................................. 25 4.2 Clean-up and HPC isolation ............................................................... 26 4.3 Derivatization of HPCs ...................................................................... 27

    4.3.1 Methylation ................................................................................. 27

  • iii

    4.3.2 Acylation with PFBCl ................................................................. 28 4.4 Analyzing biological conjugates of HPCs .......................................... 29 4.5 Instrumental analysis .......................................................................... 30

    4.5.1 GC/ECNI-MS .............................................................................. 31 4.6 Quantification of algal pigments ........................................................ 32

    5 Results and discussion ............................................................................. 33 5.1 OH-PBDE production in C. tenuicorne ............................................. 33

    5.1.1 Light and BPO, the hypothesis according to Occam .................. 33 5.1.2 BPO, and the unlikely congener pattern ..................................... 34 5.1.3 Ceramium tenuicorne and allelochemistry ................................. 38

    5.2 Concentration of OH-PBDEs over the summer season ...................... 40 5.3 OH-PBDEs in blue mussels ............................................................... 41

    5.3.1 Fatty acid conjugation of OH-PBDEs ........................................ 41 5.3.2 Conjugation of OH-PBDEs to hydrophilic moieties ................... 42

    5.4 Transport of OH-PBDEs through a Baltic Sea food chain ................. 43 5.4.1 Metabolic debromination............................................................ 44 5.4.2 Implications of the biotransformations ....................................... 47

    5.5 OH-PBDEs in grey seals .................................................................... 48 5.5.1 Exposure pattern in grey seals ................................................... 48

    5.6 Development and refinement of analytical methods for HPCs .......... 49 5.6.1 The incompatibility of diazomethane and hemolysis .................. 50 5.6.2 Extraction of coagulated seal blood ........................................... 52 5.6.3 The pros and cons of PFBCl ....................................................... 53

    6 Future perspective ................................................................................... 56 6.1 Towards a HPC monitoring program ................................................. 57

    Statement of responsibilities ...................................................................... 58 Svensk sammanfattning.............................................................................. 59 Acknowledgment ......................................................................................... 61 References .................................................................................................... 63 Appendix A

  • iv

    Abbreviations

    ADME A[D][T]P [D][T]BP BPO CA Chl a CoA DDE DDT DMF ECD ECNI [F]FA GC HELCOM HFBA[A] HPC LC LC50 LogKow LVPI [MeO-]PBDE MS NAD[+][H] OH-PBDE [OH-]PCB OXPHOS [P]CP PFB[Cl]

    Absorption, distribution, metabolism, excretion Adenosine [di][tri]phosphate [Di][Tri]bromophenol Bromoperoxidase Carboxylic acid Chlorophyll a Coenzyme A 1,1-bis(4-chlorophenyl)-2,2-dichloroethene 1,1-bis(4-chlorophenyl)-2,2,2-trichloroethane N,N-Dimethylformamide Electron capture detector Electron capture negative ionization [Free] fatty acid Gas chromatography Baltic marine environment protection commission Heptafluorobutyric acid [anhydride] Halogenated phenolic compound Liquid chromatography Lethal concentration 50 Logarithm of the octanol-water partition coefficient Liquid-vent pressure-pulsed injection [Methoxylated] polybrominated diphenyl ether Mass spectrometry Nicotinamide dinucleotide [oxidized][reduced] Hydroxylated polybrominated diphenyl ether [Hydroxylated] polychlorinated biphenyl Oxidative phosphorylation [Penta]chlorophenol Pentafluorobenzoyl [chloride]

    pKa PP PPD ROS SL SMHI TH TPP �x+c

    Negative logarithm of the acid dissociation constant Pressure-pulsed injection Photosynthetic photon density Reactive oxygen species Splitless injection Swedish meteorological and hydrological institute Thyroid hormone Thiamine pyrophosphate �xantophylls and carotenoids

  • 1

    1 Background

    The Baltic Sea, in particular the Baltic proper, is an ecosystem out of bal-ance. High inputs of nutrients to the system has resulted in eutrophication. Nutrients like phosphorous and nitrogen are essential, yet, these become harmful when released into aquatic ecosystems in excessive amounts. Via run-off from agricultural-land large amounts of nutrient has leaked, and are still leaking, into the Baltic Sea. The Baltic Sea is also the recipient of sew-age water from millions of people living in its drainage area, which adds to the amount of nutrients. However, large improvements have been made through the establishment of sewage treatment plants, and phosphates were recently prohibited in laundry- (2008) and dishwashing detergents (2011) by the Swedish government. Nutrient reduction plans have also been agreed upon within HELCOM, which includes targets for maximum allowed inputs of nutrient to the different sub-basins of the Baltic Sea (adopted by all Baltic coastal states and the EU as part of the Baltic Sea action plan in 2007, and revised in 2013) [4].

    The addition of nutrients affects the whole ecosystem, starting with primary producers. It may lead to increased primary production, as seen in the south-ern Baltic proper during the summers [5]. It may also result in shifts in the primary producer communities where certain species may increase at the cost of others, as the observed increase of filamentous macroalgae along the coast of the Baltic Sea [5]. An increase of certain algae species will succes-sively lead to increased release of potentially toxic secondary metabolites produced by these species.

    The Baltic Sea has been severely polluted by man-made toxicants, and the wildlife has suffered as a result. However, as a result of the restriction and bans of particularly polychlorinated biphenyls (PCBs) and DDT starting in the 1970s, the Baltic fauna have seen a lot of improvements in terms of pol-lution exposure and effects thereof. For example, the seals in the Baltic Sea have largely recovered from suffering severe malformation and reproduction disorders [6-8], and the white-tailed sea eagle (Haliaeetus albicilla) have been brought back from the brink of extinction [9,10].

    Yet other signs of distress are emerging. Ominous signs of an energy imbal-ance in the Baltic food web have been observed over the past decades. De-creased size and/or fat deposits has been reported in fish, marine birds and seals [11-13]. Such a large-scale effect is likely multifactorial. However, this thesis is focused on halogenated phenolic compounds (HPCs). Many HPCs possess the ability to affect the energy metabolism in exposed organisms

  • 2

    negatively, most directly via disruption of oxidative phosphorylation (OXPHOS). HPCs were once considered an environmental concern of pure anthropogenic origin. However, in recent years there has been increased interest towards the contribution of naturally produced HPCs to the overall HPC load and toxic burden in the marine fauna. Natural production of HPCs, specifically brominated phenolic compounds have been associated with a variety of organisms in the Baltic Sea including several species of algae [14,15].

    Algae have been around for millions of years, and some of the secondary metabolites of interest have at least been traced back to before the beginning of the industrial revolution [16]. However, a recent study indicated that the levels of some of these compounds, specifically some hydroxylated poly-brominated diphenyl ethers (OH-PBDEs), have increased in Baltic Sea biota over the last few decades [17]. Increased production and release of these compounds may add additional pressure to an already stressed ecosystem.

  • 3

    1.1 Aims of the thesis

    The aim of this thesis was primarily to study naturally produced HPCs in Baltic Sea biota, particularly OH-PBDEs. This included studying the levels in different compartments of the food web, the production in algae, and the transport and fate of these compounds in the ecosystem. Specific aims of the papers included in this thesis are listed below:

    Paper I: The purpose of this article was to study production of OH-PBDEs in the filamentous red algae Ceramium tenuicorne, and how this may be linked to photosynthetic pigments. The overall goal was to further our knowledge on how and why the OH-PBDEs are produced. The project was spawned following an observation that samples from a collected time series of C. tenuicorne seemingly darkened in color towards the concentration peak of OH-PBDEs in the algae.

    Paper II: The aim of this paper was to study the conjugation (esterification) of OH-PBDEs with lipophilic moieties in blue mussels (Mytilus edulis). The goals were to determine whether or not this metabolic pathway could work as a detoxification mechanism for the mussels, and what impact these types of conjugates could have when conducting exposure and environmental risk assessments that include mussels.

    Paper III: In this article, the transport of OH-PBDEs and methoxylated-PBDEs (MeO-PBDEs) through a Baltic Sea food chain was studied, from C. tenuicorne via Gammarus spp. and three-spined stickleback (Gasterosteus aculeatus) to perch (Perca fluviatilis). The purpose was to study how these compounds distribute in the food chain, and make observations on potential biotransformations that occur during the trophic transport.

    Paper IV: The aim of this paper was to develop a method for the determina-tion of OH-PBDEs in blood coagulates, then apply it to determine the OH-PBDE concentrations in grey seals (Halichoerus grypus) from different areas of the Baltic Sea.

  • 4

    2 HPC essentials

    In this chapter an introduction to HPCs is provided, including their sources to the Baltic Sea, natural production, chemical properties, transport and transformation in biota, and toxicity.

    2.1 Natural production of HPCs

    Today over 4500 naturally occurring organohalogen compounds have been documented, originating from both biogenic and abiotic sources [18]. The vast amount of sources is exceeded by the myriad of chemical structures leading to this large documented and ever growing number. Organohalogen compounds are abiotically formed in volcanos [18], and even in outer space, as observed in carbonaceous meteorites where chlorobenzoic acids have been identified [19]. Although iodine and fluorine due occur, chlorine and bromine are the most common halogen substituents in the more than 3700 organohalogen compounds produced by living organisms [20].

    Due to the accessibility of halogens in our seas and oceans, many biogenic organohalogen compounds have their source in the marine environment. As redox potential is more important than concentration in the biosynthesis of organohalogen compounds a large part of these are brominated. The com-plexity of the chemical structures among natural organohalogens stretches from the simplest haloalkanes to e.g. halogenated pyrroles, idoles, phenols, polyphenols, and glycopeptides [18]. The marine producers include every-thing from bacteria to plants, fungi and sponges, as well as marine animals like tunicates, sea slugs, soft corals, and acorn worms [18]. Phenolic struc-tures are common in the vast amount of marine orhanohalogens, and HPCs production has a large spread among marine producers [18].

    Bromophenols (BPs) and OH-PBDEs have been detected in marine bacteria, algae, and sponges all over the world [18] including the Baltic Sea [14,21]. OH-PBDEs contain two phenyl-rings bound together by an ether bridge, with one hydroxyl group (OH group) attached to one of the rings in either ortho (o), meta (m), or para (p) position, and one to nine bromine substitu-ents. General structures of BPs and OH-PBDEs are presented in Figure 2.1. Structures for all OH-PBDEs quantified in this thesis are presented in Ap-pendix A, Figure A1.

  • 5

    HPCs are also produced by terrestrial organisms, in everything from lichen and fungi, where e.g. the fungi antibiotic drosophilin A is produced [22,23], to insects and arachnids, where e.g. 2,6-dichlorophenol (2,6-DCP) is produced as a pheromone by ticks [24]. Even higher animals produce HPCs like amphibians and even humans [20], with the latter producing the iodinated thyroid hormone, thyroxine, but also chlorinated and brominated tyrosine [20]. Figure 2.2 presents some examples of naturally produced organohalogen compounds including phenolic structures.

    Figure 2.2 Examples of some naturally produced HPCs. The two first are products of the marine environment, in sponges and ascidians respectively, and the latter two are terrestrial products, in fungi and man respectively [20].

    2.1.1 Biosynthesis of bromophenols (BPs)

    Production of phenol as a precursor of BPs in e.g. algae has been suggested to proceed via decarboxylation of 4-hydroxybenzoic acid, which in turn is formed from chorismate (the precursor of e.g. aromatic amino acids) [25]. Chorismate is synthesized via the shikimate pathway, a pathway specific to microorganisms, plants, algae, and fungi, which connects the metabolism of carbohydrates to the synthesis of aromatic compounds [26]. 4-Hydroxy-benzoic acid has also been proposed to be formed as an intermediate in the biosynthesis of 2,4,6-tribromophenol (2,4,6-TBP) from L-tyrosine in the green algae Ulva lactua, in a multistep reaction where several of the suggested intermediates have been isolated [27]. 2,4-Dibromophenol (2,4-DBP) has also been shown to be synthesized directly from 4-hydroxybenzoic acid by a flavin-dependent decarboxylase-brominase present in marine bacteria [28]. Figure 2.3 shows the pathway for the biosynthesis of phenol

    Spongiadioxin A Prepolycitrin A Drosophilin A 3-bromotyrosine

    Figure 2.1 General structures forBPs and OH-PBDEs.

  • 6

    starting with phosphoenolpyruvate, and erythrose 4-phosphate. Figure 2.3 further depicts the link between the shikimate pathway and the synthesis of L-tyrosine as well as the formation of 4-hydroxybenzoic acid from L-tyrosine.

    Figure 2.3 Suggested biosynthesis routes for the production of phenol. Starting with phosphoenoylpyruvate and erythrose 4-phosphate [25,27].

    Transaminase

    Shikimate

    Chorismate

    Prephenate 4-hydroxyphenylpyruvate L-Tyrosine

    4-hydroxybenzoate

    Phenol

    Phosphoenolpyruvate+

    Erythrose 4-P4 step reaction

    3 step reactionC

    laisenrearrangem

    ent

    5 st

    ep r

    eact

    ion

  • 7

    BPs can be formed by bromoperoxidase- (BPO) catalyzed bromination of phenol. BPOs are common antioxidant enzymes in marine organisms such as algae [29]. The main function of BPOs is likely to scavenge endogenous hydrogen peroxide (H2O2) [29,30]. Reactive oxygen species (ROS) such as hydrogen peroxide are endogenously formed during the processes of photo-synthesis and respiration in plants and algae [31,32]. Hydrogen peroxide formed during these processes has in turn been observed to be available as a substrate for BPO [33]. Thus, the action of BPO may counteract hydrogen peroxide induced autotoxicity. The actual synthesis of BPs probably pro-ceeds via the BPO-catalyzed formation of hypobromous acid (HOBr) from bromide and hydrogen peroxide. The hypobromous acid, being electrophilic, will then react with phenol through an aromatic substitution reaction [25,29]. Other potential activated Br intermediates include Br2 and Br3¯ [34]. The resonance structures arising from donation of electrons from the oxygen of the phenol to the phenyl ring, renders higher electron density in ortho and para position relative to the OH group, than in meta position. This promotes electrophilic attacks at these positions (from e.g. hypobromous acid). This preference for bromination in ortho and para positions leads to a selectivity in the production of BPs. Figure 2.4 shows the formation of 2,4-DBP via a two-step reaction starting with the BPO-catalyzed formation of hypobro-mous acid.

    Figure 2.4 Simplified reaction scheme of a two-step reaction leading to the formation of 2,4-DBP from phenol, starting with an initial BPO-catalyzed formation of HOBr from H2O2 and bromide.

    2.1.2 BP dimerization and production of OH-PBDEs

    OH-PBDEs can in turn be formed by dimerization of BPs. This dimerization can also be catalyzed by peroxidases including BPO, as previously observed in vitro [35]. The dimerization is suggested to proceed through a radical reaction via the peroxidase mediated formation of bromophenoxy radicals (Figure 2.5A). The bromophenoxy radicals can react either with other phe-noxy radicals, or with non-radical BPs, as depicted in Figure 2.5B respec-tively 2.5C. The reaction in Figure 2.5B may also occur following an initial tautomerization to the o-hydroxyphenyl radical [36].

    1) 2)

  • 8

    Because of the resonance structures of the BPs, and their radicals (the latter shown in figure 2.5A), the dimerization can result in a number of structures depending on the substitution pattern of the BPs involved, including both o- and p-phenoxyphenols as well as a dihydroxylated biphenyls (the last in the case of biradical dimerization).

    OH-PBDEs could potentially also be formed by direct bromination of phe-noxyphenol. However, direct chemical bromination of o-phenoxyanisole did not yield a congener pattern recognized in nature, and hence it was discarded as a potential route for natural production of MeO-PBDEs [37].

    Figure 2.5 A) BPO-catalyzed oxidation of BPs to phenoxy radicals. B) Birad-ical dimerization of BPs (in this example two 2,4-DBP radicals). C) Dimeriza-tion reaction between a phenoxy radical and non-radical BP (in this example a 2,4-DBP radical and 2,4,6-TBP). The product in both example B) and C) is 2’-OH-BDE68.

    2.1.3 Benefit of the production

    Halometabolites are presumed to work as sinks for hydrogen peroxide [30], and at the same time they have been suggested to contribute to the survival and fitness of the producers. For example via antibiotic qualities [29,38], as defense against epiphyte settlement [39], or as grazer deterrents [38,40]. The roles and effects of different halometabolites are probably as diverse as their structures. The biosynthesis of OH-PBDEs, both with regards to mechanisms and possible effects, will be discussed further in chapter 5.1.

    A)

    B)

    C)

  • 9

    2.2 Anthropogenic sources of HPCs

    2.2.1 Intentional and accidental releases

    Chlorophenols (CPs) have been used exten-sively, particularly as fungicides for the protection of wood [41]. The intention has never been for these to spread and cause harm beyond their point of usage. However, these applications have led to a near global contamination, predominantly of pentachlo-rophenol PCP (see Figure 2.6). Today PCP is banned or restricted in most countries. In Sweden PCP was used until 1978. Up to that point (from 1956) some 2000 tons of PCP had been used in Swe-den [41]. PCP is still found in Baltic wildlife today [42].

    Chlorinated phenolic compounds have also been accidently created and re-leased into the environment through human activity, particularly via the use of chlorine gas bleaching in the paper and pulp industry [43,44]. Discharge from these facilities contain a myriad of both phenolic and neutral chlorinat-ed compounds including chlorinated phenols, catechols, guaiacols, vannilins, dibenzo-p-dioxins, and furans [43,44]. Chlorinated hydroxylated diphenyl ethers, dioxins and furans have also been associated with discharge from sawmills, paper/pulp and wood preserving facilities, together with CPs, due to the historical use of the latter as wood preservatives [45,46]. Many of these chlorinated compounds are homologous to naturally produced bromin-ated compounds. Some are directly homologous like BPs, OH-PBDEs and brominated dibenzo-p-dioxins, which have been identified in various algae, bacteria, and sponges in the Baltic Sea [3,14,15,47]. Others share important structural similarities like the large variation of brominated catecholic com-pounds isolated from algae, for example lanosol [48] (Figure 2.7, compound 2B). Figure 2.7 depicts the similarities between some chosen examples of chlorinated compounds related to wood and paper industries, and naturally produced brominated compounds.

    BPs also have an anthropogenic production. 2,4,6-TBP is the most widely produced BP with an annual production of 9500 metric ton worldwide (in 2001) [49]. 2,4,6-TBP have been produced as a wood preservative similarly to the CPs. But 2,4,6-TBP has also been produced as a flame retardant in-termediate [49]. Other brominated phenolic compounds have been applied directly as flame retardants, such as tetrabromobisphenol A, which has had an annual worldwide production during the beginning of the 21th century measured in upwards of hundred thousand metric ton [50].

    Figure 2.6 Structure of PCP.

  • 10

    Figure 2.7 Substances 1A-D are examples of compounds that have been associated with paper and wood industry. 1A, 1B, and 1D have been identi-fied in spent bleach liquors [43,44,51], 1C have been identified in soil around wood industry [45]. 2A-D are naturally produced in Baltic Sea biota [14,15].

    2.2.2 Metabolites of anthropogenic organohalogen pollutants

    HPCs can also be formed via metabolic phase I oxidation of other anthropo-genic substances, such as PCBs and polybrominated diphenyl ethers (PBDEs) from which OH-PCBs and OH-PBDEs respectively can be formed [52-54]. Thus, OH-PBDEs can be of both anthropogenic and natural origin. PCBs and PBDEs are widely spread in the environment due to their histori-cal extensive usage. The levels of PCBs and PBDEs are decreasing in the Baltic Sea biota as a result of their bans and/or restrictions [11]. However, their high lipid solubility, stability, and capacity to biomagnify makes them still fairly abundant in fatty species and top predators [11,55,56].

    The source to the OH-PBDEs in an organism can often be elucidated from the congener pattern. Specifically, by observing the position of the OH group among the detected OH-PBDEs. The most abundant naturally produced OH-PBDEs in the Baltic Sea have the OH group in ortho position in relation to the ether bridge. Metabolically derived OH-PBDEs on the other hand have been observed to have the OH group primarily in para or meta position [57,58]. But, o-OH-PBDEs do also occur among PBDE metabolites, for example 6-OH-BDE47 have been indicated as a metabolite of BDE47 [59,60].

    1A 1B 1C 1D

    2A 2B 2C 2D

  • 11

    2.3 HPCs, physicochemical properties and kinetics

    2.3.1 Physicochemical properties

    HPCs are weak acids. They are fairly lipophilic in their protonated form, while they are more water soluble in their dissociated state. Both the oc-tanol-water partition coefficient (Kow) and the acid dissociation constant (Ka) can stretch over several orders of magnitude between different HPCs. Both factors are dependent on the substitution pattern, specifically the degree of halogenation. More halogen generally means more acidic, but also more lipophilic. The predicted pKa and logKow values for the seven, tetra- to hexa-brominated, OH-PBDEs quanti-fied in this thesis stretches from 4.7 to 6.8, and from 5.9 to 7.4 respectively (calculated using ACD/Labs v.11.02). The highest logKow and lowest pKa is held by the only hexa-brominated congener (see Table 2.1). The pKa and logKow can largely influence the environmental fate of HPCs.

    2.3.2 Adsorption, distribution, metabolism, excretion (ADME)

    The pKa of the OH-PBDEs studied in this thesis (Table 2.1) are all either close to or below neutral (physiological) pH. Thus, chemical properties and partition coefficients, such as logKow, determined on the protonated state are of little environmental relevance. Due to the comparably high pH in the Bal-tic Sea (~7.9-8.3, east of Gotland, according to SMHI) absorption of these compounds through passive diffusion by e.g. fish becomes less relevant than other exposure routes (e.g. via feeding). Other factors such as vapor pressure and the water-air partition coefficient are also affected by the dissociation, which means that it influences the abiotic transport of these substances as well.

    In the case of HPCs with pKa values below ambient pH the spatial arrange-ment of the substituents may be more important with regards to ADME than the degree of halogenation. For example, the substitution pattern can largely affect affinity for blood proteins etc., which co-governs the retention of HPCs in higher organisms. HPCs, have previously been observed to have selective retention in blood [61], and they have been observed to association

    Compound pKa logKow

    2'-OH-BDE68 6.6 6.7

    6-OH-BDE47 6.8 5.9

    6-OH-BDE90 5.65 6.8

    6-OH-BDE99 5.2 6.6

    2-OH-BDE123 5.65 7.2

    6-OH-BDE85 6.2 6.6

    6-OH-BDE137 4.7 7.4

    Table 2.1 Predicted pKa and logKow for the main OH-PBDEsanalyzed in this thesis. Calculatedusing ACD/ Labs v.11.02.

  • 12

    to blood proteins such as transthyretin [62], serum albumin [63], and some lipoproteins [64].

    Other important factors for the retention and distribution of HPCs in the food web include biotransformation. HPCs have a suitable functional group for phase II metabolism. However, the capacity to metabolize phenolic com-pounds varies between species, even within the same family, as observed for different species of fish [65]. Gold fish (Carassius auratus) has been ob-served to excrete PCP via the gills unconjugated, but most of the excreted PCP was conjugated with sulfate [66]. Sulfate conjugates can be excreted via the gills or kidneys, while glucuronic acid conjugates are excreted almost exclusively via the bile [66,67]. CPs have been shown to have a half-life of eight days or less in whitefish, related to the possibility of direct conjugation (phase II metabolism) and excretion [68]. OH-PBDEs being much larger than CPs have longer half-lives in fish, up to twenty days for 2’-OH-BDE68 in common carp (Cyprinus carpio) [69], which may be related to an excre-tion more restricted to the biliary route.

    Phase II metabolism is generally aimed to facilitate excretion. However, some metabolic routes can also generate more lipophilic compounds, which may increase the retention within the organism. O-methylation of HPCs resulting in the corresponding anisoles has been shown to occur in various strains of bacteria [70]. HPCs can also be conjugated with fatty acids via fatty acid acyltransferases (discussed more in chapter 5.3.1). For example, PCP has previously been observed to be conjugated with palmitic acid, both in vitro and in vivo [71,72]. Some examples of phase II metabolic products of HPCs are shown in Figure 2.8.

    HPCs also undergo biotransformations associated with phase I metabolism, such as enzymatic oxidation to quinones (via oxidative dehalogenation) [36,73], and reductive dehalogenation [74]. Cleavage of the ether bond be-tween the rings may also occur for OH-PBDEs. It has even been shown to be the dominant biotransformation pathway in pig liver microsomes, especially for lower brominated congeners [75].

  • 13

    Figure 2.8 Some examples of phase II metabolites of HPCs, in this case 2,4,6-TBP, including (from left to right) a sulfate conjugate, a glucuronic acid conjugate, an acyl conjugate (R=undefined hydrocarbon chain), and a me-thyl ether (anisole).

    2.4 HPC toxicity, with focus on OH-PBDEs

    HPCs such as OH-PBDEs may elicit toxicity through several modes of ac-tion, for example through hormonal disturbance and disruption of oxidative phosphorylation (OXPHOS). OH-PBDEs have also been shown to be cyto-toxic [76], genotoxic [77], and neurotoxic [78]. OH-PBDEs also cause em-bryo toxicity in fish, which as observed in developing zebrafish (Danio re-rio), can result in malformations [79,80]. Examples of such malformations include spinal curvature, which has been confirmed as a sensitive biomarker for exposure to 6-OH-BDE47 during development [79]. 6-OH-BDE47 is also known to cause fin malformations [81].

    Spinal curvature has also been observed on sculpin (Myoxocephalus quadri-cornis) living in drainage areas of wood and paper factories [82]. These mal-formations were associated with highly chlorinated phenolic compounds by multivariate data analysis [82]. Other cartilage/skeletal deformations ob-served on the fish living in these areas included damage fins and lesions [82].

    2.4.1 Endocrine disruption

    HPCs can disrupt the endocrine system in a number of ways depending on their structure. OH-PBDEs may for example act as both estrogens and anti-estrogens depending on the substitution pattern. Specifically, lower bromin-ated OH-PBDEs tend to act agonistically while higher brominated OH-PBDEs act antagonistically [83]. p-OH-PBDEs have been shown to bind to both the thyroid receptor [84] and the thyroxine transport protein transthyret-in [85]. BPs have also been shown compete with thyroxine for transthyretin [85,86], while 6-OH-BDE47 in comparison display only mild competitive

  • 14

    potency [86]. The high potency for p-OH-PBDEs to compete with thyroxine is likely related to their structural similari-ties, as thyroxine contains a p-OH-tetra-iododiphenyl ether structure (see Figure 2.9). In contrast o-OH-PBDEs, especially those lacking an adjacent bromine to the OH group such as 6-OH-BDE47, seem to be favored in binding to the thyroid hor-mone transport protein, thyroxine-binding globulin [87].

    2.4.2 OXPHOS disruption

    The majority of adenosine triphosphate (ATP), the body’s main energy cur-rency, is formed by OXPHOS in the mitochondria of eukaryotic cells. OXPHOS connects nutrient catabolism with respiration in the phosphoryla-tion of adenosine diphosphate (ADP). This is achieved by a series of redox reaction, in which electrons are transported from NADH and succinate to molecular oxygen via the electron transport chain. The chain is made up by a number of complexes bound in the inner mitochondrial membrane (see Fig-ure 2.10), complex I, III and IV uses the energy from the redox reactions to dislocate protons across the membrane creating a proton gradient (pH gradi-ent) and separation of charge (��). The energy stored in this formed electro-chemical potential is in turn utilized by the ATP synthase to phosphorylate ADP via backflow of protons (Figure 2.10).

    The OXPHOS process can be disrupted in a number of ways. For example, 6-OH-BDE47 has been shown to inhibit complex II (succinate dehydrogen-ase) in the electron transport chain [80]. Similarly, the natural product and pesticide/piscicide rotenone inhibits complex I (NADH dehydrogenase). Complex III (ubiquinol:cytochrome c oxidoreductase) can be inhibited by the herbicide 2,4-dichlorophenoxyactic acid, cyanide inhibits complex IV (cytochrome c oxidase), and several mycotoxins inhibit the ATP synthase [88].

    OXPHOS may also be disrupted (uncoupled) by substances that can shuttle protons over the inner mitochondrial membrane, which thus diminishes the proton gradient powering OXPHOS [88]. These so called protonophoric uncouplers, including a number of OH-PBDEs [89], can achieve this due to the roughly 0.6 points pH difference between the intermembrane space and the matrix. This difference allows e.g. a OH-PBDE with a low enough pKa to occur mainly in its dissociated form at physiological pH, to accept a

    Figure 2.9 Chemical structureof thyroxine, 3,3’,5,5’-tetraiodo-L-thyronine.

  • 15

    proton in the more proton dense intermembrane space, diffuse through the membrane and release it in the matrix [88]. Due to the membrane potential associated with the proton gradient, the anions can then migrate back over the membrane and the cycle starts all over. This last step of the cycle is the rate-determining step since it requires a lot of energy to push the anions through the fatty membrane. Because of this, phenols with bulky ortho groups that can shield (occlude) the charge become powerful uncouplers [88,90]. The same is true for phenols with the ability to form dimers during the migration back over the membrane, i.e. phenols that can form anion hydrogen bonding between a protonated and dissociated phenol (as suggested for 2,4-dinitrophenol) [88,90]. During the process of uncoupling the energy destined for incorporation in ATP will be released as heat causing thermogenesis. Proton leak over the membrane can also follow as a result of substances that damage the structural integrity of the membrane or alter the membrane fluidity, as achieved by general anaesthetics such as halothane [88].

    While exposure to high levels of OXPHOS disrupting chemicals can lead to acute toxicity, prolonged exposure to lower doses can result in lowered en-ergy production, and may cause wasting as well as hyperthermia [88].

    Figure 2.10 Simple schematic of OXPHOS including the complexes of the electron transport chain and the ATP synthase in the inner mitochondrial membrane. The yellow arrow depicts the electron flow (from right to left in the figure).

    Matrix

    Inter-membrane

    space

    ADP+PiATP

    O2H2O H+ H+H+

    H+

    ���pH

    CoQCoQCyt C

    IIIIIIIV

    AT

    P sy

    nt.

  • 16

    2.5 Ecosystem concerns and the larger perspective

    2.5.1 Xenobiotic exposure and metabolic disruption

    To understand the concerns and ultimately the adverse role that OH-PBDEs may play on an ecosystem like the Baltic Sea, it is necessary to put the ef-fects of these compounds into a larger perspective. The OH-PBDEs are only one of several factors that may adversely affect the Baltic wildlife. Other toxins and pollutants in the Baltic Sea share toxicological mechanisms and endpoints with OH-PBDEs, which can lead to additive or even synergistic effects. For example, several other chemical groups besides OH-PBDEs target OXPHOS, such as perfluorinated acids and sulfonamides (uncouplers) [91,92], as well as diarylamine pesticides such as the rodenticide brome-thalin and the fungicide fluazinam (uncouplers) [93]. Exposure to perfluori-nated acids (and sulfonic acids) can also effect the lipid metabolism of an organism [94].

    Hydrogen sulfide released from anoxic sediment via the action of e.g. prote-olytic bacteria [95], can also affect OXPHOS via inhibition of cytochrome c oxidase (complex IV in Figure 2.10). Hydrogen sulfide has been observed to be highly toxic to several species of fish [96]. The spread of hypoxic sedi-ment in the Baltic Sea [97] is of large concern, and both perfluorinated sub-stances and OH-PBDEs have increased in Baltic Sea biota over the last 30 years [17].

    Furthermore, the metabolism in an organism can be affected by several mechanisms of which OXPHOS disruption is only one. For example, OH-PBDEs also affect the thyroid hormone system [86,87], which indirectly can affect the metabolism. These hormones have a regulatory role in the metabo-lism, specifically on basal metabolism, lipolysis/lipogenesis, and gluconeo-genesis [98] (see Figure 2.11). Exposure to thyroid hormone disrupting sub-stances may also lead to hyperthyriodism, as indicated in cats [99]. Many other environmental toxins target the endocrine system as well, both natural and anthropogenic. These include everything from brominated flame retard-ants to pesticides, organophosphorus compound, and constituents of anti-fouling paints used on boats (for an extensive review see [100]).

  • 17

    2.5.2 Co-affecting factors, thiamine deficiency

    There seem to be a wide spread thiamine deficiency in the Baltic fauna. Thi-amine or rather thiamine pyrophosphate (TPP) is an essential cofactor for many metabolic enzymes (specific sites are indicated in Figure 2.11). Thia-mine deficiency can have very striking effects, such as paralysis observed in birds in the Baltic Sea [101]. Thiamine deficiency among Baltic wildlife is not a new phenomenon. The M74 syndrome related to the reproduction of salmon (Salmo salar), which has been monitored since 1974, has for exam-ple been associated with thiamine deficiency [102,103]. However, the phy-toplankton community in the Baltic Sea may have a lower capability to sup-ply thiamine to the food web today than in the early 1980s following a shift in the phytoplankton community [104]. Specifically, since dinoflagellates (that have less species producing thiamine) have increased while diatoms (that have more species producing thiamine) have decreased [104]. Thiamine levels in an organism can also be affected by intake of anti-thiamine com-pounds [105], and even a simple molecule like catechol has been shown to degrade thiamine in vitro [106].

    Thiamine deficiency can affect the metabolism and cause energy imbalance in an organism long before the levels drop below the border of lethality [101]. Thus, with affected energy production as the toxic endpoint, the effect concentration of an OXPHOS disrupting compound is likely to become low-er in an organism suffering from thiamine deficiency. Figure 2.11 displays a simplified schematic overview of the metabolism (specifically the catabo-lism), in which the enzymes using thiamine are indicated together with the metabolic sites targeted (directly and indirectly) by OH-PBDEs.

  • 18

    Figure 2.11 Simplified schematic of the metabolism with indicated sites af-fected by TPP deficiency, major sites affected by TH disturbance, and sites affected by OH-PBDEs directly.

    Oxaloacetat

    Succinyl-CoA

    �-Ketoglutarate

    Citric acidcycle

    Urea cycle

    Acetyl-CoA

    PyruvateLactate

    Phosphoenolpyruvate

    Glyceraldehyde 3-P

    Fructose 6-P

    Glucose 6-P

    Glucose

    Glycogen Ribulose 5-P

    Xylulose 5-P

    Ribose 5-PTransketolase

    Pentosephosphatepathway

    Glu

    cone

    ogen

    esis

    Glycogenesis

    Urea

    Proteins

    FAAs

    �-oxidation

    FFAs

    Lipids

    Lipolysis

    Proteolysis

    NH3Pro

    tein

    syn

    thes

    is

    Glycolysis

    �-Keto acids

    Branched-chain�-Keto acids

    IIQ

    c IIII

    IVA

    ADP

    ATP

    O2H2O

    e-

    e-OXPHOS

    Pyruvate-DH

    �-ketoglutarate-DH

    Protongradient

    Ox.

    Non-Ox.

    OH-PBDEs

    TH

    TH

    TH

    Direct targets of OH-PBDEs

    Anabolic routes

    Enzyme using TPP as cofactorXX

    CoA

    DH

    FAAs

    FFAs

    Ox.

    Non-Ox.

    P

    TH

    = Coenzyme A

    = Dehydrogenase

    = Free amino acids

    = Free fatty acids

    = Oxidative phase

    = Non-oxidative phase

    = Phosphate

    = Thyroid hormone regulated

  • 19

    3 Samples and sample collection

    3.1 Species studied in this thesis

    3.1.1 Ceramium tenuicorne

    The filamentous macroalga C. tenuicorne (Figure 3.1) (Kützing) Waern (Ceramiaceae) is one of the most common and widely distrib-uted (along the saline gradient) red algae species in the Baltic Sea. It can grow on an array of substrates, including rocks and other algae, as well as in loose drifting algal mats [107,108]. It is capable of both sexual and vegetative reproduction. However, the occurrence of sexual reproduction is strongly reduced along the saline gradient in the Baltic Sea [108]. The wide distri-bution of C. tenuicorne in the Bal-tic Sea is likely related to local adaptation and not so much to a general tolerance for different sa-linity levels [109]. Ceramium tenuicorne belongs to a group of algae gener-ally considered to be favored by nutrient enrichment [109-111]. Fast grow-ing opportunistic species with short life-cycles are generally favored over perennial species, such as Fucus vesiculosus, in the presence of excess nutri-ents [110,111]. However, the effect of nutrient enrichment on C. tenuicorne differs along the saline gradient of the Baltic Sea, with positive effects seen in the Baltic proper but not in the Gulf of Bothnia [109].

    The nutrient enrichment in the Baltic Sea has resulted in an increase of fila-mentous algae along the coast [5]. At the same time the coverage of the structurally important keystone specie F. vesiculosus, which form the basis of the most species-rich biotope in the Baltic Sea, has decreased over the last decades [112,113]. In the archipelago of Östergötland in the Baltic proper the coverage of C. tenuicorne increased dramatically between years 2007 and 2011 [114]. Ceramium tenuicorne has previously been suggested to be a producer of OH-PBDEs [3,14,115], and it has been proven to produce BPs [3].

    Figure 3.1 Photography of Ceramium tenuicorne collected for Paper I.

  • 20

    3.1.2 Blue mussel

    Blue mussels (Mytilus edulis, Figure 3.2) constitute the largest biomass in the Baltic Sea. They make up more than 80% of the total animal biomass of hard-bottoms down to below 30 m in the Baltic proper [116]. The domi-nance by blue mussels in the Baltic proper, follows from having few pred-ators and competitors for space [116]. Mytilus edulis is a marine species that has successfully adapted to the brack-ish waters of the Baltic Sea. However, the blue mussels in the Baltic Sea grow slower, and reach smaller sizes than their counterparts in marine waters. This is presumably related to saline de-pendent metabolic differences [117]. In the north of the Baltic Sea, in the Bothian bay, the salinity is too low to be habitable for M. edulis. Blue mus-sels are efficient filter feeders, and a single blue mussel may filter through several liters of water per hour [118]. Through the combined efforts of all blue mussels in the Baltic Sea it has been estimated that the entire volume Baltic Sea may be turned over by the blue mussels in one year [118].

    The blue mussels although still dominant, have decreased in number in the northern Baltic proper since the beginning of the 1990s [119]. In the south-ern Baltic proper, in Hanö bay, the blue mussels also exhibit poor condition with decreased soft-tissue to shell ratio [120].

    3.1.3 Gammarus spp.

    Gammarus spp. (Figure 3.3) are amph-ipod crustaceans. There are hundreds of species belonging to the Gammari-dae genus both fresh- and marine water species. Likewise, in the Baltic Sea, several species are represented. Includ-ed among the more common species found in the Baltic Sea are: G. duebeni, G. locusta, G. oceanicus, G. salinus, and G. zaddachi. The gammarids in the Baltic Sea are both grazers and scav-engers feeding both on algae and dead degrading material, including animalic

    Figure 3.2 Photography of a Baltic blue mussel (Mytilus edulis).

    Figure 3.3 Photography of aGammarus sp. collected forPaper III.

  • 21

    material. The gammarids analyzed in Paper III may have contained several species. The individuals chosen for analysis were generally of an average size close to 1 cm, and up to 30 individuals were pooled for a single sample.

    3.1.4 Three-spined stickleback

    The three-spined stickleback (Gas-terosteus aculeatus, Figure 3.4, from here on only referred to as stickle-back) is one of the most common fish species in the Baltic Sea, and increas-ingly so as it is a species that has seen a notable population increased in recent years [121]. Although their increase may be perceived as being beneficial for stickleback feeding piscivores such as perch, it may actu-ally be the opposite. A large stickle-back population can affect the re-cruitment of perch and northern pike (Esox lucius) by a large predation on their eggs [122]. Genetic studies on stickleback have indicated that some populations in the Baltic Sea have un-dergone rapid evolution brought on by human activities. For example genetic changes have been observed in populations inhabiting polluted sites close to pulp mills, and populations dwelling in areas highly affected by eutrophica-tion [123].

    Stickleback can survive and breed both in marine and fresh water environ-ments. The population turn-over of stickleback proceeds rapidly, they reach sexual maturity at ages 1-2 and seldom live longer than 3 years. In high-summer after spawning a lot of sticklebacks die, suggestively due to ex-haustment and depleted energy reserves from spawning [124]. However, sticklebacks in the Baltic Sea are also largely affected by parasites, specifi-cally by the tapeworm Schistocephalus solidus, which was observed among the individuals analyzed in Paper III as well. The stickleback act as an in-termediate host in the worm’s life-cycle. The combination of the effort in-vested in spawning, the energy drained by the parasite infestation, and expo-sure to energy-metabolism disrupting compounds such as OH-PBDEs, may lead to increased mortality of old individuals.

    Figure 3.4 Photography of a three-spined stickleback (Gasterosteusaculeatus) from the Baltic proper.

  • 22

    3.1.5 Perch

    The perch (Perca fluviatilis, Figure 3.5) population in the Baltic Sea has decreased in recent years in contrast to that of stickleback. The decrease is potentially related to poor recruitment. In 80% of investigated outer archipe-lagic sites in the Baltic Sea the recruitment was considered weak or com-pletely absent [122]. The poor recruitment has been suggested to be related to starvation of the larvae due to a shortage of zooplankton [125]. The perch is a fresh-water specie that can also inhabit brackish waters. Spawning takes place in late spring, and compared to the stickleback the population turn-over is slow. The perch reach sexual maturity at around 2-6 years, and can live for as long as 20 years. This may potentially make the perch population more vulnerable to effects of chronic exposure to toxicants than the stickle-back population. The perch grow significantly larger than stickleback and can weigh up to 2-3 kg and reach length of up to 40 cm. Young perch with a size around 15 cm were chosen for the trophic transfer study conducted in Paper III.

    Figure 3.5 Photography of a perch (Perca fluviatilis) from the Baltic proper.

    3.1.6 Grey seal

    The Grey seal (Halichoerus grypus, Figure 3.6) population in the Baltic Sea is recovering after having been hunted for a century, and suffered from poor recruitment (low fertility rates) for several decades associated with environ-mental pollutants [7]. Specifically the prevalence of uteri occlusions and skull bone lesions has gone down in association with the decreasing trends of DDT (and its degradation products) and PCBs [6]. Today the grey seal population in the Baltic Sea is increasing by 7-8% annually [8]. However, the prevalence of colonic ulcers increased during the period between 1977

  • 23

    and 1996 [6], and the blubber thickness has been decreasing during the start of the 21th century [12].

    The grey seals are divided into two subspecies, with H. grypus macroryn-chus being the one found in the Baltic Sea. The grey seals can live upwards of 40 years and the males, being larger than the females, can grow up to 2.5 m in length and weigh up to 300 kg [126]. The grey seals reach sexual ma-turity at an age of around 4-6 years. Breeding takes place in late spring, and the young are born early spring in the following year. The pups suckle for about 3 weeks, over this period the female loses 40-50% of its body weight [126]. The grey seals in the Baltic Sea feed mainly on fish with herring (Clupea harengus membras) being the most common food source, although many species are included in their diet [126]. The seals feeding preference lead them to compete for resources with commercial fishing, and they are often found foraging in fishing equipment, such as nets, where they may get entangled and drown.

    Figure 3.6 Photography of a grey seal (Halichoerus grypus) kept in captivity.

    3.2 Sample collection

    The samples analyzed in Paper I and Paper III were collected at Nämdö Island, indicated on the map in Figure 3.7, in the summer of 2011 and 2013 respectively. All samples were collected from a south facing bay area of approximately 1000 m2. Ceramium tenuicorne samples were rinsed in seawater, and all associated visible organisms were removed. Stickleback samples were collected using landing nets, and perch using gill nets (Paper III). The stomach content of the sticklebacks and perch were removed and

  • 24

    excised through ocular observation, to determine feeding preferences, and to define trophic links in the food chain. 8 to 12 individual fish were pooled into a single sample. All samples were homogenized using a blender with stainless steel blades, and approximately 10 g of the homogenates were taken for analysis whenever possible.

    Blue mussels were collected at the marine laboratory at Askö (see Figure 3.7) during the summer of 2008 (Pa-per II). The samples were collected using a bottom scraper. During each sampling occasion, roughly one liter of blue mussels (containing hundreds of individuals) was collected. The mussels were removed from the shells and homogenized using a stainless steel blender. Four 10 g samples from each homogenate were taken for anal-yses.

    All seal blood samples analyzed in Paper IV were donated by the Swe-dish museum of natural history. The blood (partially coagulated) had been collected during autopsies of deceased grey seals sent in to the museum after being found dead, e.g. in fishing equipment. Biological measurements including size, gender and health sta-tus (e.g. blubber thickness) were rec-orded for all individuals. The samples were obtained from seals originating from south to north along the Swedish Baltic Sea coast (see approximate locations in Figure 3.7), which had died between the years 2005 and 2007. All prepared samples were fro-zen and stored at -20 C until analyses were conducted.

    = I/III= II= IV

    Paper

    Figure 3.7 Map over the SwedishBaltic Sea coast with indicatedsampling locations. Grey triangleindicates Nämdö Island where thesamples for Paper I and Paper IIIwere collected. The white squareshows the site where the musselsfor Paper II were collected, theAskö field station. Black circlesindicate the general areas wherethe grey seals for Paper IV werefound dead.

  • 25

    4 Analytical methods

    4.1 Extraction

    Organohalogen compounds in general have often been extracted from tissue samples using classical lipid extraction methods such as the Bligh and Dyer [127], its forerunner the Folch method [128], the later developments of these procedures [129,130], and the Jensen method [131]. Additionally, methods such as soxhlet extraction, microwave-assisted extraction, and accelerated solvent extraction has also been used [132,133].

    Different types of solvent extraction methods have also been developed for bodily fluids [134-136]. For small samples of simple matrices such as serum and plasma, dilution followed by direct application on solid phase extraction columns has been applied [137]. Accelerated solvent extraction has also been applied on freeze-dried blood samples [138]. This subject has been extensively reviewed elsewhere for further reading [139].

    4.1.1 The lineage of the Jensen methods

    The so-called “original Jensen” method (named after its inventor, prof. Sören Jensen) saw the light of day in 1972 [131]. Following the spawned interest in persistent lipophilic chlorinated contaminants, this method was developed to provide an alternative extraction procedure for accurate deter-mination of lipid content, and quantitative extraction of chlorinated contami-nants, which did not include the use of chlorinated solvents. Chlorinated solvent were otherwise standard in the methods of the time [127,128]. How-ever, in 2003 it was determined that the Jensen method had a limitation re-garding the extraction of phospholipids. Thus further developments were made, which resulted in a new method (“new Jensen”) [140]. The “new Jen-sen” method applied isopropanol instead of acetone in combination with diethyl ether and hexane to increase the efficiency to extract phospholipids. This new version also included some further developments previously done to the original method [53].

    In 2009 a final modification of the method was made, in which the filtration step, to separate the liquid extract from undissolved protein residue, was omitted in favor of centrifugation [1]. The newest method, appropriately called “Jensen centrifugation”, was developed with the aim to create a simpler and faster method. Although the same solvents that had been applied earlier was used [140], the ratios were changed to maintain extraction efficiency with fewer steps. The final method was also evaluated for

  • 26

    determination of HCPs and alkyl phenols in marine organisms, and included clean-up and derivatization procedures to target these compounds [1].

    The “Jensen centrifugation” extraction method [1] was applied in the studies presented in Paper II and Paper III. The algae samples analyzed in Paper I and Paper III respectively, were extracted using the so-called “new Jensen” [140]. Filtration is more suitable for removing the algal (C. tenuicorne) resi-due as the fine particulates can be difficult to condense properly to a firm pellet using centrifugation. In Paper IV a new method is presented for blood samples, which will be discussed in chapter 5.6.

    4.2 Clean-up and HPC isolation

    pH partitioning was applied to separate HPCs from neutral compounds by taking advantage of the acidity of the HPCs. The partitions were conducted using potassium hydroxide in 50% ethanol. The ethanol works as an organic modifier that increases solubility of the dissociated HPCs in the alkaline phase. It also counteracts precipitation of free fatty acid (FFA) alkali salts, which if plenty can largely affect the outcome of the separation. After the separation, the alkaline phase was acidified and the HPCs back-extracted into organic solvents. Other techniques that can be successfully applied to isolate HPCs from neutral compounds include ion-exchange chromatog-raphy.

    Acid-catalyzed esterification was applied to neutralize FFAs (Paper II and Paper IV). FFAs have previously been observed to cause poor yield during derivatization of HPCs with pentafluorobenzoyl chlorid (PFBCl) [1], which was used in the analyses presented in Paper II and Paper IV. However, FFAs are not a problem when using diazomethane to derivatize HPCs, as done in Paper I and Paper III, as long as large enough excess is used (as the FFAs rapidly react with the reagent). FFA esterification has been con-ducted within this thesis using sulfuric acid (H2SO4), hydrochloric acid (HCl), and boron trifluoride (BF3) in methanol. The reaction was done over 1 h at 70°C regardless of which of the three catalysts that was used.

    During analysis of blue mussels an additional clean-up step was applied prior to derivatization in the form of a deactivated (5% water) silica gel col-umn (0.5 g) [1] (Paper II). This step was added to remove some polar com-ponents (pigments etc.) found in the mussel extract. Treatment with concen-trated sulfuric acid was applied after derivatization, as a final clean-up step prior to instrumental analysis (Paper I-IV). Sulfuric acid destroys matrix components through oxidation. It can also be used to remove saturated fats

  • 27

    through protonation of the carbonyl oxygen in the ester and formation of a sulfuric acid soluble complex, (R-COH-OR’)+HSO4-. The removal of satu-rated fatty acid esters is important when analyzing neutral compounds such as MeO-PBDEs, as done in Paper III. When analyzing tissues containing large amounts of lipids, bulk removal techniques such as gel permeation chromatography (GPC) [133,141], and acetonitrile partitioning [142] are often applied following the extraction.

    4.3 Derivatization of HPCs

    Several derivatization reagents have been used for the determination of HPCs. Derivatization is used to enhance the chromatographic properties of the acidic HPCs when conducting analysis using gas chromatography (GC). Although GC has been the most popular analytical instrument for HPCs for a long time, nowadays more and more work is being conducted with liquid chromatography/mass-spectrometry (LC/MS). LC/MS can be applied for both non-derivatized HPCs [138], and HPCs derivatized to increase sensitiv-ity [137]. Some examples of reagents used for derivatization of HPCs prior to GC analysis include methylating (alkylating), various acylating, as well as silylating compounds.

    Within the work conducted for this thesis two main derivatization techniques have been utilized, methylation, and acylation with PFBCl.

    4.3.1 Methylation

    Methylation of HPCs generates stable derivatives that can be treated harshly during sample clean-up and stored over long periods of time. However, re-garding OH-PBDEs it is important to separate them from possible MeO-PBDEs originally occurring in the samples before derivatizaton. Methylation is often conducted using quite hazardous reagents. In the analyses presented in Paper I and Paper III diazomethane was used for this reaction, which is extremely hazardous being acutely toxic, carcinogenic, teratogenic, and ex-plosive. Diazomethane was synthesized from N-methyl-N-nitroso-p-toluene-sulfonamid (Diazald) using a slightly modified version of the procedure de-scribed by Vogel [143]. Special permission to synthesize, store, and handle the diazomethane had to be acquired from the Swedish Work Environment Authority. However, once prepared, this reagent is easy to use. An excess of reagent is added to the samples and once the reaction is completed, the ex-cess is evaporated. The reaction generally leaves little residual interferences that need to be removed before instrumental analysis.

  • 28

    Iodomethane was also used for methylation of OH-PBDEs, as an alternative to diazomethane (Paper II). The method was developed based on the proce-dure described by Bergman and Wachtmeister [144]. The derivatization with iodomethane generates high yields but is more laborious than the procedure with diazomethane. Iodomethane is also an alkylating agent that should be handled with great care. However, it is neither explosive nor a gas at room temperature, and therefore requires less precaution than diazomethane. Methylation of HPCs has also previously been conducted using trimethylsi-lyldiazomethane, as in the analysis of mammalian blood [145].

    4.3.2 Acylation with PFBCl

    The acylation with PFBCl is conducted in a biphasic system with an organic solvent and an aqueous carbonate-based buffer containing a phase transfer catalyst (tetrabutylammonium) [1]. The HPCs are ionized and partition into the buffer where they interact with the catalyst to form a lipophilic salt. The formed salt then distributes to the organic phase and there reacts with the PFBCl reagent. The result is an addition of a fluorinated moiety to the ana-lyte, which tends to enhance the signal when using electron capture detection (ECD) or electron capture negative ionization (ECNI) MS, due to a higher electron negativity and ionization efficiency. However, for already large HPCs the addition of a large PFB moiety may cause poor chromatography, counteracting an increase in signal intensity by peak broadening, and poor injection efficiency.

    The reaction with PFBCl is easy to perform and fast (2 min at ambient tem-perature). However, the method is sensitive to matrix components, and hence requires proper clean-up procedures before used on complex samples [1]. Removal of excess reagent can also be difficult, and fluorinated interferences can deteriorate the performance of GC/ECD analyses. The PFB acyl deriva-tives of HPCs are also less stable over time compared with methyl deriva-tives. The use of PFBCl will be discussed more in chapter 5.6.3.

    Other acylating agents successfully adapted to OH-PBDEs within the work of this thesis include acetic acid anhydride and heptafluorobutyric acid an-hydride (HFBAA). The first reagent is easy to use and of low cost, but the formed derivatives do not tolerate sulfuric acid to the same extent as the PFB or HFBA esters. The HFBAA derivatization suffers from similar problems as the PFBCl method, especially those related to the clean-up after derivati-zation. However, the HFBA derivatives exhibit better chromatographic properties. Acetic acid anhydride and HFBAA were used for comparisons and during method developments, but they were not used in any of the in-cluded quantitative studies.

  • 29

    4.4 Analyzing biological conjugates of HPCs

    OH-PBDEs occurring as lipid soluble (fatty acid (FA)) esters were released by conducting transesterification of the neutral fraction, isolated after the pH partitioning (Paper II). In contrast to hydrolysis, transesterification does not generate FFAs. Instead, FAs are converted to (in this case) methyl esters (Figure 4.1). Two different methods were used, one acid-catalyzed method (10% BF3 in methanol, 70°C over night) (see Figure 4.1), and one base-catalyzed method (0.25 M sodium methoxide in methanol, 70°C for 10 min). These methods were evaluated using OH-PBDE FA esters synthesized in-house (Paper II).

    Aqueous soluble OH-PBDE conjugates were on the other hand hydrolyzed. These conjugates partition into the aqueous (water/isopropanol) phase fol-lowing the extraction. Additional acid was thus added to this phase followed by reflux at ambient pressure for 1 h (Paper II). As no proper standards were available to evaluate the hydrolysis of aqueous soluble conjugates, the proper time and acid strength were determined using extracted fish viscera, and human urine (kindly provided “in-house” by professor emeritus Sören Jensen). The acid-catalyzed method was also compared with a base-catalyzed method to ensure similar results. A schematic over the analytical procedures applied in Papers I-IV is shown in Figure 4.2.

    Figure 4.1 Reaction scheme for transesterification of a OH-PBDE acyl con-jugate, in this case 6-OH-BDE47, using boron trifluoride in methanol (R=undefined hydrocarbon chain).

  • 30

    Figure 4.2 Flow chart of the analytical procedure applied within this thesis. Blue sections are only included in Paper II, red in Paper III and green in Paper IV. Extractions are conducted according to [140] in Paper I and for algae in Paper III, and according to [1] in Paper II and Paper III. In Paper IV a new extraction procedure is described. Derivatization is conducted using diazomethane in Paper I and Paper III, and using iodomethan in Paper II. PFBCl is used in Paper II and Paper IV according to [1]. Lip.=Lipid soluble conjugates, Aq.=Aqeous soluble conjugates, Free=free phenols, CA=carb-oxylic acid. Modified from a scheme in Paper II.

    4.5 Instrumental analysis

    GC/ECD has probably, historically, been the most used setup for separation and detection of halogenated contaminants in general. Although ECD was scarcely used for quantification in the included papers (only used for DDE in Paper III, and to verify results), it was extensively used for method devel-opment and screenings conducted within the work of this thesis. Most of the

    Lip.+ Free

    Extraction

    pH partitioning Hydrolysis

    Transesterification

    pH partitioning SiO2 column

    Derivatization

    EsterificationCA-groups

    H2SO4 treatment

    Aq.

    FreeLip.

    GC/MS

    MeO-PBDEs

    II, IV

    I, III II

    IV

    III

    II

  • 31

    HPC determinations conducted today apply MS for detection and quantifica-tion, either coupled to GC or LC. GC/ECNI-MS is a very common technique concerning brominated substances like OH-PBDEs.

    4.5.1 GC/ECNI-MS

    GC/ECNI-MS was used for quantification in all studies. ECNI utilizes a buffer gas to create low energy (thermal) electrons. Methane was used in the analyses conducted in Paper I, III and IV, and ammonia in Paper II. The thermal electrons are “captured” by electronegative groups in a molecule, e.g. halogens, which produce negative ions. Thus, ECNI is a fairly selective ionization technique for halogenated compounds, and the ionization efficien-cy is generally high (given that enough number of halogens are present in the molecule).

    Bromine atoms attached to carbons have a low energetic unoccupied molec-ular orbital suitable for accepting an electron in ECNI mode [146]. Bromin-ated organic compounds tend to undergo dissociative electron capture upon ECNI, often via cleavage of the carbon-bromine bond, in which a bromide ion and a radical are formed, according to the reaction scheme:

    C-Br + e¯ → [C-Br]·¯ → C· + Br¯

    This dissociative electron capture process also occurs for methylated OH-PBDEs (MeO-PBDEs). This makes selective ion monitoring (SIM) of the bromide isotope ions (m/z 79, 81) a both selective and sensitive method for their detection. However, the low relative abundance or nearly total absence of molecular ion signals in the ECNI spectra (due to low preference for asso-ciative electron capture) of brominated compounds can be a problem when conducting qualitative analyses.

    PFBCl derivatization of OH-PBDEs introduces an additional weak bond in the structure (in the form of an ester bond), which readily fragments in ECNI mode to produce phenoxy-phenolate ions (¯O-PBDEs, see Figure 4.3). Similar fragmentation occurs also for the HFBA derivatives, but in this case the negative charge tends to end up on the HFBA group, which means that there is no structural information gained related to the

    Figure 4.3 Ionized PFB derivativeof a OH-PBDE using GC/ECNI-MS.Scission occurs in the ester bond tocreate a phenoxyphenolate ion.

  • 32

    analyte. However, the HFBAA derivatization can be used quite successfully in combination with SIM analyses. When using acetic acid anhydride derivatization, the OH-PBDEs fragment to produce mostly bromide ions, despite the introduction of an ester bond within the structure.

    4.6 Quantification of algal pigments

    Pigments were extracted from C. tenu-icorne (Paper I) using ultrasonic as-sisted extraction with cold N,N-dimethylformamide (DMF), similarly to a previously described method [147]. The extracts were diluted with DMF, and filtered before spectropho-tometric analysis. The concentration of chlorophyll a (Chl a) was calculated using Lambert-Beer’s law (Equation 1) with the measurement at absorbance maxima in the red spectrum region (at 664 nm). The absorption coefficient (α) for Chl a in DMF has previously been determined to 90.41 L/g×cm [148]. Figure 4.4 shows an absorbance spectrum of a C. tenuicorne extract in DMF.

    Equation 1 l

    ACa ���

    664

    The concentration of�xantophylls and carotenoids (�x+c) were calculated according to a previously derived equation for �x+c in DMF, with absorb-ance measurement at 480 nm (Equation 2) [148]. Equation 2 accounts for the absorbance contribution of both Chl a (Ca in the equation) and Chl b (Cb) at 480 nm. However, C. tenuicorne does not contain Chl b [149], hence making the Cb factor redundant.

    Equation 2 245

    02.5289.01000 480 bacx

    CCAC

    ���

    400 500 600 700

    Abso

    rban

    ce

    Wavelength (nm)

    Figure 4.4 Absorbance spectrum ofa C. tenuicorne extract in DMF (datafrom Paper I).

  • 34

    Hydrogen peroxide formed during photosynthesis can be used by BPO to brominate phenol (see chapter 2.1.1). Production of BPs has even been suggested as the primary mean by which two alga species of the Rodomelaceae family scavenge hydrogen peroxide [30]. As OH-PBDEs can be formed by BPO-catalyzed dimerization of BPs [35], it seems reasonable that the produc-tion rate of OH-PBDEs could be linked to photosynthetic rate in C. tenuicorne. Photosynthetic rate in turn is correlated to the density of photosynthetic pigments and incom-ing irradiance in algae [150]. If the levels of pigments increase without a corresponding decrease in irradi-ance, as observed in Paper I (see Figure 5.1), then the light absorption that drives the photosynthesis will increase in the alga. (Figure 5.2 dis-plays the connection between photo-synthesis and production of OH-PBDEs in a simplified flow chart that include the reactions shown in Figure 2.3-2.5).

    The observed correlations between pigments and OH-PBDEs thus seem to support the general assumption that OH-PBDEs are biosynthesized via BPO-catalyzed dimerization of BPs. However, the OH-PBDE congener pattern in the C. tenuicorne samples analyzed in Paper I did not fully agree with this assumption, which will be discussed in the next chapter.

    5.1.2 BPO, and the unlikely congener pattern

    BPO can undoubtedly catalyze the reactions required to form of OH-PBDEs (see Figure 2.4 and 2.5) as proven by in vitro experiments [35]. However as stated by Lin and co-workers, no OH-PBDE congeners that have been iden-tified in nature were formed in the in vitro experiments with BPO [35].

    Photosynthesis

    BPO

    H2O2

    BPO

    H2O2

    Carbohydratemetabolism

    Photosyntheticpigments

    Irradiance

    Shikimate

    Figure 5.2 Simplified flow chartdisplaying the link between irra-diance, pigmentation, photosyn-thesis, and the production of OH-PBDEs via the action of BPO.

    CO2

    ATP

  • 35

    BPO does not seem to contribute too much selectivity in the synthesis of OH-PBDEs. It oxidizes the reactants, but neither the bromination reaction (Figure 2.4) nor the dimerization reaction (Figure 2.5) seem to be spatially coordinated on the surface of the BPO enzyme, which means that selectivity will be determined only by the resonance of phenol (see chapter 2.1.1). Fur-ther, BPO does not catalyze dimerization of 2,4-DBP, in vitro, in the pres-ence of bromide and hydrogen peroxide [35]. Instead 2,4-DBP is converted to 2,4,6-TBP [35]. Once 2,4,6-TBP is formed the bromination sequence seem to reach a thermodynamic sink. From this point homo-dimerization commences via the formation of 2,4,6-tribromophenoxy radicals, which leads to the production of both the corresponding o- and p-phenoxyphenol, at a statistical distribution of 2:1 [35].

    The conger pattern in C. tenuicorne is dominated by 6-OH-BDE85 and 6-OH-BDE137, generally followed by 6-OH-BDE47. Together these three congeners made up more than 87% of the �OH-PBDE concentration in the samples analyzed in Paper I, and two of these were the only OH-PBDEs that were detected in C. tenuicorne cultivated under experimental conditions [3]. Further, all dominant OH-PBDE congeners in C. tenuicorne are o-(2,4-dibromophenoxy)BPs. To achieve this pattern via dimerization of BPs, an enzyme that selectively oxidizes 2,4-DBP to a phenoxy radical, as well as spatially coordinates the dimerization (to exclusively form o-phenoxy-phenols) would be needed. Further, to be formed via this pathway each of the three dominant congeners would require a BP (aside from 2,4-DBP) that has not been successfully produced by BPO-catalyzed bromination of phenol [35,151-153]. Similar congener patterns have also been observed in other macroalga species from the Baltic Sea (see Table 5.1)

    Thus far only vanadium BPO has been evaluated with regards to production of OH-PBDEs, and it seems unlikely that this enzyme should be solely re-sponsible for production of OH-PBDEs from phenol, in C. tenuicorne. How-ever, FeHeme BPO has also been detected in algae, and this enzyme is known to catalyze classical peroxidase reaction including oxidation of phe-nols to phenoxy radicals [29]. The type of BPO present in C. tenucorne is unknown but other species of the Ceramium genus have been observed to contain at least vanadium BPO [29]. The synthesis of OH-PBDEs via dimer-ization of BPs has also been catalyzed by other enzymes in vitro, such as laccase and bacterial cytochrome P450 [28,154]. However, only one of the major OH-PBDEs found in C. tenuicorne from the Baltic Sea (2’-OH-BDE68) was formed among the produced OH-PBDEs in the experiments with these enzymes [28,154].

  • 36

    It is possible that OH-PBDEs are not formed by dimerization of BPs at all in C. tenuicorne, and other pathways should not be discarded. OH-PBDE pro-ducers belonging to other kingdoms or domains may of course produce dif-ferent OH-PBDE congener patterns than the Baltic macroalgae because of different biosynthetic pathways. The differences in the MeO-PBDE pattern between algae and marine sponges has previously been highlighted [146].

  • 37

    Pape

    r I

    Pape

    r II

    ID

    ata

    aqur

    ied

    from

    Löf

    stra

    nd 2

    011

    [14]

    Red

    alg

    aeC

    ompo

    und

    Cer

    amiu

    mte

    nuic

    orne

    Red

    alg

    aeG

    reen

    alg

    aeBr

    own

    alga

    e2,

    4-D

    BP

    0.24

    0.23

    Poly

    siph

    onia

    Furc

    ella

    riaC

    lado

    phor

    aEn

    tero

    mor

    pha

    Pila

    yella

    2,4,

    6-T

    BP

    0.61

    0.94

    fuco

    ides

    lum

    bric

    alis

    glom

    erat

    ain

    test

    inal

    islit

    tora

    lis2'

    -OH

    -BD

    E68

    0.14

    0.15

    0.86

    0.21

    0.19

    0.90

    0.10

    6-O

    H-B

    DE

    470.

    190.

    253.

    71.

    20.

    340.

    580.

    07

    6-O

    H-B

    DE

    900.

    060.

    10N

    D0.

    140.

    200.

    360.

    04

    6-O

    H-B

    DE

    990.

    130.

    19N

    D0.

    270.

    230.

    380.

    09

    2-O

    H-B

    DE

    123

    0.06

    0.06

    ND

    0.11

    0.04

    0.05

    0.01

    6-O

    H-B

    DE

    851.

    51.

    12.

    91.

    21.

    30.

    780.

    706-

    OH

    -BD

    E13

    71.

    01.

    01.

    01.

    01.

    01.

    01.

    0

    Tabl

    e5.

    1R

    elat

    ive

    mol

    arco

    ncen

    tratio

    npa

    ttern

    inm

    acro

    alga

    efro

    mth

    eBa

    ltic

    Sea,

    rela

    ted

    toth

    eco

    ncen

    tratio

    nof

    6-O

    H-B

    DE1

    37(s

    etto

    1.0)

    ,mos

    tabu

    ndan

    tcon

    gene

    rsin

    dica

    ted

    inbo

    ld.

  • 38

    5.1.3 Ceramium tenuicorne and allelochemistry

    6-OH-BDE137, has been observed to be bioactive against Gram-positive bacteria as well as fungi, and crustaceans [155]. In fact, the determined LC50 for 6-OH-BDE137 (1.4 μM) in the brine shrimp (Artemia salina) (48h) le-thality test is orders of magnitude lower than those reported for aromatic carbamate insecticides in the corresponding bioassay (29.4 and 963 μM for carbaryl and propoxur respectively) [155,156]. The high bioactivity towards organisms that potentially could cause harm to C. tenuicorne suggests that the OH-PBDEs may serve as allelochemical defense agents (allomones) for C. tenuicorne.

    In the data from Paper III it can be observed that the occurrence of OH-PBDEs and MeO-PBDEs in the amphipod Gammarus spp. can be attributed to grazing on C. tenuicorne. Both the �OH-PBDE and �MeO-PBDE con-centration trends between the two species were significantly correlated over the summer season with 95% confidence levels (see Figure 5.3A and B). Yet the ratio between the concentration of �OH-PBDEs in Gammarus spp. and the concentration in C. tenuicorne decreased in the beginning of the summer (Figure 5.3A). The same was observed for the �MeO-PBDEs concentration (Figure 5.3B).

    It was surprising to see the �MeO-PBDEs concentration ratio between the species decrease over time, and with the same general trend as for the OH-PBDEs (p=0.049, R2=0.57, given by regression analysis between the trends). It was expected that this ratio would increase over time because of the bioac-cumulation of MeO-PBDEs in Gammarus spp. A potential reason to the decreased bioaccumulation of MeO-PBDEs over time can be that the gam-marids changed their feeding preference over this period towards a lower intake of C. tenuicorne. Decreased grazing on C. tenuicorne could also ex-plain why the �OH-PBDEs concentration ratio between the species behaved in a similar manner as the �MeO-PBDE ratio. Interestingly the decrease in these ratios coincided with the increase of OH-PBDEs in the algae (see Fig-ure 5.3C). This could indicate that these compounds deter grazing by Gam-marus spp. Other brominated phenolic compounds produced by the red algae Odonthalia corymbifera have been shown to exhibit feeding-deterrent activi-ty [40]. However, foraging preference by Gammarus spp. may also have changed following a potential seasonal variation in the algae species compo-sition over the studied period, as observed during the sample collection for Paper I.

  • 41

    5.3 OH-PBDEs in blue mussels

    During analysis of Baltic blue mussels, high concentrations of HPC esters were observed in the lipid extract (prof. Sören Jensen, personal communica-tion). It was hypothesized that the esterification of HPCs could work as a detoxification mechanism for mussels. If HPCs could be sequestered as fatty acid esters during periods of intense exposure it may be advantageous for the blue mussels as it would minimize the activity of the endocrine- and OXPHOS disrupting OH-PBDEs during the exposure peak in the summer.

    5.3.1 Fatty acid conjugation of OH-PBDEs

    OH-PBDEs may be converted to FA esters in blue mussels via the action of acyltransferases. These enzymes are pronounced in mollusks such as blue mussels as they are utilized to regulate steroid homeostasis [157,158]. It was initially hypothesized that proper acyltransferases would be induced in the blue mussel during the exposure peak of OH-PBDEs, which would lead to increased formation of FA esters. The corresponding effect has previously been observed on mussels exposed to estradiol, during which the percentage of esterified estradiol increased with increasing exposure [159]. However, the expected increase in the percentage of esterified OH-PBDEs was not observed during the exposure peak in the mussels analyzed in Paper II (see Figure 5.5).

    The highest percentage of esterified OH-PBDEs in the mussels was observed in May (roughly 50%, see Figure 5.5). The concentration of FA esters then increased to reach the highest observed level in June, i.e. when the exposure to OH-PBDEs was at its highest. However, the percentage of OH-PBDEs that were conjugated as FA esters was at its lowest at this time (less than 10%, see Figure 5.5). The decrease in the percentage of OH-PBDEs that were conjugated as FA esters during the peak of exposure may be an indica-tion that the expression of the metabolic enzyme was not, or only slightly, induced. From this observation, it can be concluded that this phenomenon is not a detoxification mechanism, or at least it works very poorly as such. The conjugation of OH-PBDEs likely occurs because of low substrate specificity of a steroid acyltransferase in the mussels (Paper II).

    The high levels of lipophilic conjugates observed in the May sample (see Figure 5.5) may be the result of an earlier concentration peak of OH-PBDEs in the mussels, occurring around March-April. Such peak could have left visible remnants in the form of these conjugates due to their comparably (to the free OH-PBDEs at least) long depuration time. Similar enrichment of FA esters, in relation to free OH-PBDEs, can be observed between the June and

  • 42

    August samples (see Figure 5.5). Because of the possibility for enrichment of these conjugates, FA esters should at least be considered during exposure assessment and mass balance studies of HPCs that include mussels. These conjugates may also be a potential tool in monitoring of HPCs as it enables the remnants of concentration spikes to be studied well after their occurrence (Paper II).

    Figure 5.5 Concentrations of �6OH-PBDEs, occurring in their free form (Free, n=4), as lipid soluble (acyl) conjugates (Lip., n=4), and aqueous solu-ble conjugates (Aq., n=3) respectively, in blue mussels (data from Paper II).

    5.3.2 Conjugation of OH-PBDEs to hydrophilic moieties

    The occurrence of aqueous soluble conjugates of OH-PBDEs was also stud-ied in the mussels (Paper II). This type of metabolism has previously been observed to be less pronounced in mollusks than in fish [160]. It was ob-served that aqueous soluble conjugates of OH-PBDEs generally occurred in much lower concentration than the FA esters (see Figure 5.5). However, the turn-over rate of these conjugates seem to be higher than that of the FA es-ters, which means that this could still be a more active metabolic pathway with regards to OH-PBDEs in mussels. The levels of aqueous soluble conju-gates (presented in Figure 5.5) may be somewhat underestimated as no prop-er surrogate standard was available for these conjugates during the analyses (Paper II).

    0

    4

    8

    12

    080528 080625 080815 081008

    [�O

    H-P

    BDEs

    ]

    13

    44

    75

    (nm

    ol/k

    g w

    w)

    Lip. Aq. Free

  • 45

    In the samples from the 25th of July and 5th of August, 2013, where all inves-tigated congeners were